Biology › Section G · Chromosomes and DNA › Topic G.4
Basics of DNA replication
The double helix was more than a pretty structure - its two complementary strands whispered how DNA must copy itself. But three rival models fit the geometry, and it took heavy nitrogen, an ultracentrifuge, and two patient generations of E. coli to prove which one nature actually uses.
The structure suggests the copy
Watson and Crick's model implied a copying mechanism. The two strands of the double helix are held together only by hydrogen bonds between complementary base pairs - A with T, G with C. Unzip the helix and each strand carries the full information of the sequence, because each base predicts its partner. So each parental strand can serve as a template: line up new nucleotides by the pairing rules, link them, and a complementary daughter strand is born. Two helixes where there was one - and each daughter cell of a division inherits one. What the structure could not say was how the old and new DNA end up distributed among the products.
Three rival models
Three mechanisms were proposed. In conservative replication, the parental helix stays entirely intact and the two new strands pair with each other - one molecule all old, one all new. In semi-conservative replication, the two parental strands separate, and each builds its own new partner - every daughter molecule is a hybrid of one old strand and one new. In dispersive replication, both daughter molecules are patchworks, with old and new segments interspersed along both strands. All three are consistent with the double helix; distinguishing them required an experiment that could tell old DNA from new.
The trick: heavy nitrogen
Matthew Meselson and Franklin Stahl found a way to label DNA by density. Nitrogen is a component of every nucleotide base, and E. coli grown for many generations in medium containing the heavy isotope nitrogen-15 incorporates it into all of its DNA. Switch the culture to the normal light isotope, nitrogen-14, and every newly made strand is light instead. Old DNA and new DNA now differ in weight - and old from new is exactly what the rival models disagreed about.
Spinning DNA into bands
To read those weights, Meselson and Stahl used density-gradient centrifugation. DNA is loaded into a cesium chloride solution and spun at 50,000 to 60,000 rpm in an ultracentrifuge. The gradient pushes heavy molecules down and buoyancy pushes them up until each molecule settles at the depth matching its own density, forming a visible band. Heavy, nitrogen-15 DNA bands lower than light, nitrogen-14 DNA; a molecule mixing both isotopes bands in between. The band pattern in the tube is a direct census of what is old, what is new, and how they are combined.
Generation one: a single intermediate band
Grow E. coli in heavy nitrogen, then shift the culture to light medium and let the cells divide once. Isolate the DNA and spin it. Meselson and Stahl observed a single band at the intermediate position - not one heavy and one light band, but one band in between. That immediately kills the conservative model, which predicts two separate bands: an intact heavy parental molecule and an all-light new one. Both remaining models survive, though - both semi-conservative and dispersive put every molecule in a half-heavy, half-light state after one generation.
Generation two: two bands settle it
Let the culture divide once more and spin again. Now two bands appear: one at the intermediate position and one at the light position, in roughly equal amounts. The dispersive model dies here. If old and new DNA were interspersed along every strand, every molecule would be a bit more than half light, and the single band would drift gradually lighter - never splitting into two. Only the semi-conservative model predicts a clean split: the heavy parental strands pair with new light partners (intermediate molecules), while the light strands made in generation one pair with new light strands (all-light molecules). Two bands, exactly as seen - replication is semi-conservative.
Why semi-conservative matters
The verdict means every replicated DNA molecule is half old, half new: one conserved parental strand, one freshly synthesized strand. The parental information is literally kept - 'semi-conservative' means the original sequence encoded in each parental strand is conserved in the daughter molecules. That conserved strand is a built-in reference copy, which matters for the next chapter of this story: proofreading and repair enzymes compare the new strand against the old one, so the hybrid design of each molecule is the foundation of replication fidelity. It also guarantees that when the chromosome copies separate into two daughter cells, each cell receives the same sequence as its parent had.
From hint to machinery
The elegant unzip-and-copy picture is true as far as it goes, but replication turns out to be far more complex than the Watson-Crick sketch. The helix does not simply fall open: enzymes unwind it, proteins hold it open, primers start each new strand, and polymerases that can only build in one direction force one strand into short fragments. The structure of DNA explains how the information is copied; the machinery of the cell explains how it happens fast and accurately - the subject of the next topic.
Worked example: predict generation three
Under semi-conservative replication, what band pattern appears after a third generation in light medium? Track the heavy strands. They never pair with each other, so every heavy strand sits in an intermediate molecule - and each generation their fraction is halved as the DNA pool doubles. After generation two the ratio was one intermediate to one light; after generation three, intermediate molecules are 1/4 of the total and light molecules 3/4. Spin the tube and you see two bands, the intermediate one a quarter as strong as before relative to the light one - a thinning heavy echo, never disappearing entirely, never splitting.
Worked example: diagnosing a rival universe
Suppose instead the generation-one DNA had formed two separate bands - one heavy, one light. Which model would that support? Conservative replication: the parental molecule stays intact and heavy, while the new pair of strands forms its own light molecule. Then in generation two, conservative predicts the heavy band persists undiminished while the light band doubles - old and new never mix. And if instead the observed pattern were a single band that merely crept lighter each generation, the verdict would be dispersive. One experiment, three signatures: the band pattern after each generation fingerprints the model - which is exactly why Meselson and Stahl's design is considered among the most beautiful in biology.
Common mistakes to avoid
First, thinking one generation was enough - after one round, semi-conservative and dispersive look identical; it takes the second generation to separate them. Second, misreading the intermediate band as a mixture of two DNA types - it is a single molecular species, each molecule half heavy. Third, forgetting the conservative model predicts two bands after generation one - observing one intermediate band is precisely what eliminated it. Fourth, calling heavy nitrogen radioactive - it is a stable isotope, separated by density, not detected by radiation. Fifth, assuming 'semi-conservative' means half the DNA is conserved - each molecule conserves one entire strand, so all the information is kept. Sixth, expecting the heavy band to vanish in generation two - heavy strands persist inside intermediate molecules, halving in proportion each generation.
Key takeaways
DNA's complementary, hydrogen-bonded strands suggested that each parental strand templates a new partner - but three models fit: conservative (parental molecule stays intact), semi-conservative (each daughter molecule one old strand, one new), and dispersive (old and new interspersed). Meselson and Stahl labeled DNA by growing E. coli in heavy nitrogen-15, then shifted it to light nitrogen-14 and separated the DNA by density in cesium chloride ultracentrifugation. Generation one showed a single intermediate band, ruling out conservative; generation two showed two bands - intermediate and light - ruling out dispersive and proving replication semi-conservative. Every replicated molecule conserves one parental strand as a reference for fidelity, and each generation halves the heavy-strand fraction while the heavy echo never splits or disappears.
Flashcards
Flashcard
Parental strand whose sequence guides a new complementary strand
Test your understanding
Question 1 of 5
After ONE generation of E. coli growth in light (14N) medium following growth in heavy (15N) medium, Meselson and Stahl observed a single band at the intermediate density. Which model does this eliminate?
Question 2 of 5
Under semi-conservative replication, what band pattern is expected after TWO generations in light medium?
Question 3 of 5
Why did Meselson and Stahl grow E. coli in nitrogen-15 before switching to nitrogen-14?
Question 4 of 5
What exactly does 'semi-conservative' mean for a replicated DNA molecule?
Question 5 of 5
Under semi-conservative replication, what fraction of the DNA is in intermediate-density (hybrid) molecules after the THIRD generation in light medium?
